Rates, Predictors and Mortality Impact of Losses from HIV Care in South Africa

描述（由候选人提供）：拟建的5年发展期将用于研究和培训活动，使Fox博士具备成为独立研究者的技能。Fox博士的长期职业目标是：1)发展评估大型临床数据集的专业知识；2)对艾滋病毒治疗方案的影响进行分析；3)开发新的分析方法来评估艾滋病毒项目对临床和社会结果的影响。这一培训将包括正式的课堂学习，开展一个研究项目，以及艾滋病毒、流行病学和数学建模专家的研究指导。培训的重点是：1)全面了解艾滋病毒管理；2)专门的流行病学方法；3)实践研究实施技能。主要研究项目是对非洲最大的艾滋病毒治疗中心之一（超过1.1万名接受抗逆转录病毒治疗的患者）——南非约翰内斯堡的Themba Lethu诊所（TLC）——的辍学生进行研究。尽管2004年在南非大规模推广了抗逆转录病毒疗法，但许多艾滋病毒检测呈阳性的患者从未开始接受抗逆转录病毒疗法，许多开始接受治疗的患者退出了治疗。关于患者在连续治疗过程中何处丢失的信息对抗逆转录病毒治疗项目的成功至关重要。本研究的目的是：1)：测量TLC损失的比率和预测因素；2)使用南非国家死亡登记处验证TLC死亡率数据；3)估计抗逆转录病毒治疗降低死亡率的效果；4)利用数学模型估计提高保留率的策略对死亡率的影响。我们将进行一项纵向队列研究（以共同导师Sanne开展的TLC队列为基础），评估三种人群中患者的退学率：1)hiv阳性患者；2)开始抗逆转录病毒治疗前艾滋病毒治疗的患者；3)接受抗逆转录病毒治疗的患者。TLC目前使用的Therapy Edge（tm）数据库将用于监测第2组和第3组。我们将受试者纳入第一组，并将使用数据库评估每个人群的退出情况。我们将对所有失去随访的患者的死亡率数据进行验证研究，并使用验证数据评估ART的有效性，使用边际结构模型比较ART前和ART死亡率。

DESCRIPTION (provided by candidate): This proposed 5 year development period will be used for research and training activities that will give Dr. Fox the skills to become an independent investigator. Dr. Fox's long-term career goals are: 1) to develop expertise in evaluating large clinical datasets; 2) to generate analyses of the impact of HIV treatment programs; and 3) to develop novel analytic approaches to estimating the impact of HIV programs on clinical and social outcomes. This training will include formal class study, carrying out a research project, and research mentorship from experts in HIV, epidemiology and mathematical modeling. The training will focus on: 1) a thorough understanding HIV management; 2) specialized epidemiologic methods; and 3) practical research implementation skills. The primary research project is a study of drop-outs from one of the largest HIV treatment sites in Africa (more than 11,000 patients on antiretroviral therapy), the Themba Lethu Clinic (TLC) in Johannesburg, South Africa. Despite the large scale roll-out of antiretroviral therapy (ART) in 2004 in South Africa, many patients who test HIV-positive never start ART and many who start drop out of care. Information on where in the continuum of care patients are being lost is of vital importance to the success of ART programs. The aims of this study are: 1): to measure the rates and predictors of losses from TLC; 2) to validate TLC mortality data using South Africa's National Death Registry; 3) to estimate the effectiveness of ART on mortality; and 4) to estimate the impact of strategies to increase retention on mortality using mathematical modeling. We will conduct a longitudinal cohort study (built on the TLC cohort run by co-mentor Sanne) assessing drop out rates among patients in 3 populations: 1) patients testing HIV-positive; 2) patients initiating pre-ART HIV care; and 3) patients on ART. The Therapy Edge(tm) database which is currently being used at TLC will be used to monitor groups 2 and 3. We will enroll subjects in group 1 and will use the database to assess drop-outs from each population. We will conduct a validation study of mortality data for all patients lost to follow-up, and use validated data to assess the effectiveness of ART comparing pre-ART and ART mortality rates using marginal structural models. RELEVANCE: To best utilize resources for maintaining patients in HIV care, detailed information is needed on when patients are lost, how we can identify them and what can be done to increase retention. The results of this study will give a complete picture of drop out and mortality from HIV care, what effect treatment programs are having on mortality, and where limited resources for retaining patients could have the greatest impact.